Does real-time inflation data lead the CPI? The Lead Test — 91 prints graded, every input audited

Two claims hide in "leads CPI". The first — that a real-time gauge's reading on the eve of a print is a better guess of the print than the previous print — is false for every Nowflation variant: since 2019 the previous print misses the next one by +0.33pp on average and the Nowflation Gauge by +0.51pp (91 prints; CPI Tracker +0.43pp, PCE-weighted +0.55pp, Cost of Living +3.69pp). The second — that the gauge's reading today carries information about CPI months ahead that CPI's own history does not — is the testable one: since 2019 the Gauge's monthly YoY correlates 0.75 with official CPI six months later against 0.74 for CPI predicting itself over the same 82 months, clearing persistence from 4 months out; in a joint regression of CPI six months ahead on both, the Gauge's t-statistic is 1.98 and CPI's own lag -0.02 (R² 0.568 on the Gauge vs 0.546 on CPI itself, n 82). Since July 2023 — a flat stretch of CPI with little variance to explain — the six-month figures are 0.36 vs 0.23 on 28 months (the same ordering, on a sample too small to lean on). Last print: Jul 2026 CPI 3.4% (released 2026-08-12); the Gauge read 2.84% the evening before (gap -0.56pp), the CPI Tracker 3.28%, the previous print was 3.5%.

Correlation with official CPI k months ahead — 2019-01 to today (paired months: 89 at k = 0, 78 at k = 12)

Since July 2023 (35 paired months at k = 0): Gauge k0 0.90, k1 0.66, k2 0.35, k3 0.07, k4 -0.02, k5 0.06, k6 0.36, k7 0.42, k8 0.23, k9 -0.01, k10 -0.25, k11 -0.48, k12 -0.64; CPI's own persistence k0 1.00, k1 0.78, k2 0.43, k3 0.10, k4 -0.07, k5 -0.05, k6 0.23, k7 0.47, k8 0.38, k9 0.21, k10 -0.07, k11 -0.41, k12 -0.72.

Every live input against the BLS subindex it feeds — lead-lag on monthly YoY (Pearson r)

The rent lead: the tenure-weighted all-tenant rent model (chain-linked Zillow ZORI / Apartment List new-lease blend, 6% monthly turnover over 36 months) leads CPI rent of primary residence by 6 months at r 0.97 (0.78 unshifted; 0.85 for CPI rent's own persistence over the same horizon); the raw new-lease blend peaks at a 13-month lead (r 0.87). Shifted forward 6 months, the model's latest reading of 1.09% YoY (Aug 2026) is where it says CPI rent (2.87% in Jul 2026) is headed by Feb 2027 — a reading, not a graded forecast.

The graded table — each variant on the eve of the release vs the print, last 24 prints

Print-eve summary since 2019 (91 prints): The Gauge MAE +0.51pp, r 0.965; CPI Tracker MAE +0.43pp, r 0.971; PCE-weighted MAE +0.55pp, r 0.973; Cost of Living MAE +3.69pp, r 0.922; previous print MAE +0.33pp, r 0.980. Last 12 prints: The Gauge +0.50pp, CPI Tracker +0.26pp, PCE-weighted +0.42pp, Cost of Living +2.31pp, previous print +0.29pp. CPI-Tracker parity (the plumbing check, last 12 prints): MAE +0.26pp, mean gap -0.10pp, r 0.821. Next CPI release: 2026-09-11.

Method: Print eve: each variant's YoY on the calendar day before every CPI release since 2019-01 (reference month = the month before the release; the annual seasonal-factor revision that shares February with the January print is dropped), graded against the print as BLS quotes it — the NSA CPI-U 12-month change at one decimal — with the previous print as the naive benchmark. Lead: the monthly average of each variant's daily YoY correlated with official CPI YoY k months later, k = 0–12, scored on the same months as two naive benchmarks — CPI's own persistence corr(CPI_t, CPI_t+k), and the real-time naive corr(CPI_t−1, CPI_t+k), the last print actually published when the gauge month closed; regressions of CPI(t+k) on the gauge, on CPI itself, and on both, with the joint t-statistic. Components: Pearson r on monthly YoY between each live input and the BLS subindex it feeds, the input shifted forward k months (rents 0–18, pump price 0–3, Manheim 0–6, farm gate 0–9); the naive columns are the subindex's correlation with its own value k months later and with its last published month. New-lease rent = chain-linked ZORI/Apartment List blend (weights 0.5/0.3 renormalized); all-tenant rent = tenure-weighted average of that blend with 6% monthly turnover over 36 months. Latest vintage of every series; no smoothing, no window search, no parameter tuned to the result. Nothing here is tuned to look good; the naive benchmark is on every chart. Related: The Gauge, Exposed, Why the Gap?, Scoreboard, methodology.

Explore: dashboard · methodology · status · heatcheck · my inflation · vs bls · jobs · housing · farm to shelf · data · scoreboard · gap · gdp · oil · oil vs cpi · gauge · pipeline · natgas · reserves · stockpiles · about · privacy · contact · analysis · expectations · activity · labor · conditions · income · trade · surveys · today · forecast · compare · embeds · money · fiscal · stress · rates · affordability · matrix · indicators · mortgage · calculator · cart · states · counters · recession · commodities · energy · countries · macro · calendar · liquidity · ratios · housing market · revisions · century · nowflation difference · nowflation terms · nowflation how to · supercore · credit · growth · real wages · jobs preview · fx · metros · prediction markets · heating season · thanksgiving · fed watch · nowflation vs truflation · lead test · eggs · gas · coffee · ground beef · electricity · milk · bread · bacon · chicken · chocolate · butter · cheese · flour · rice · potatoes · tomatoes · bananas · oranges · cola · i bonds · tax brackets · all 1,629 series